Policy Research J WORKING PAPERS intemnatlonal Trade Intemational Economics Department The World Bank September 1993 WPS 1174 Asian Trade Barriers Against Primary and Processed Commodities Raed Safadi and Alexander Yeats Tariff escalation to protect domesdc industries against more efficient producers is not limited to industrial countries. Protec- tion of domestic industries is also common in Asian developing countries and in intra-Asian trade. PwlicyRowinaWogPapwdi netobtfinf npofw a andpeowagetfeexchangcnfideunamgBa ffand adobbaundi_eboWaddBTmepatsepmsboa dbythoRmearMviyStffycutfenamaofU hoisefelct set b.adbhdtothsWodd Ban. i.Board ifDiwcu,iwmangni.oruay of hsmcmbrcouitri Polley Resouch| 4~~~~~~~~~~~~~~~~~~ International Trade WPS 1174 This paper-a productofthe Intemational Trade Division, Intemational Economics Departnent-is part of a larger effort in the departnment to analyze and predict structual changes in trade and to identify factors affecting developing countries' exports. Copies of the paper are available fice from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Jean Jacobson, room S7-037, extension 33710 (September 1993, 29 pages). Many developing counties are being encouraged against imports of processed goods is built into to shift toward incmased processing and exports trade barrier escalation among Asian countries of domestically produced natural-resource-based and should be addrrssed in regional initiatives to products now exported in primary form. But in liberalize intra-Asian trade barriers. many major import markets, the structure of tariffs and nontariff barriers militate against such Safadi and Yeats make three recommenda- efforts. tions for dealing with escalation issues in muld- lateral negotiations: Zero or low tariffs are generally applied to * Japan and, to a lesser extent, the Republic of industrial countries' imports of primary (unproc- Korea are the key to successful negotiations on essed) comnodities; duties increase, or "esca- these issues, as they have a far greater import late," as the level of processing or fabrication bias against processed commodities than do all increases. Tariff escalation produces a trade bias other countries with which Safadi and Yeats against processed goods. compare them. That is, Japanese and Korean trade barriers incorporate far more escalation In the past, such trade banier escalation has than do trade barriers in other countrties studied. been attributed chiefly to industrial countries. Safadi and Yeats examined the structure of * Disproportionately high cuts in trade restrictions in Asian countries and found that barriers for unprocessed commodities are not the most Asian countries' tariffs incorporated more solution, as they would increase effective escalation than do tariffs in industrial countries. protection for processed goods. Apparently tariff escalation is also often rein- forced by nontaiiff barriers on processed goods, * Any approach to trade liberalization should although supporting data for this finding are less deal with both tariffs and nontariiff barrizrs, to firm. ensure that a reduction in one type of restriction is not offset by a futher tightening in the other. This issue should not be viewed as a North- Several Asian countries apply both types of South issue, contend Safadi and Yeats. A bias restricdons to commodity imports. ThePolicy ReseaWorkin PaperSiesdisseminates thebmdinPsofwolkunder wayin theBis AnobjsidveofioesnC is to got thee findings out quiclcly. even if presenations are less dfian fuDfy polised. The i'mdings. mtupretalions. md coclcusionls in thiem papers do nlot necessarily resent of ficial Bank policy. Produced by the Policy Reseah Di_ehmon Center Asian Trade Barriers Against Primary and Processed Commodities by Raed Safadi and Alexander Yeats Economists, International Trade Divibion Intemational Economics Department World Bank, Washington, DC 20433 I. Introduction Theoretical models of the development process and actual plans for industrialization often assign a key role to trade policy measures. Moreover, in addition to the maintenance of steady growth in exports and accompanying increases in foreign exchange earnings, developing countries repeatedly stress the need to reduce their dependence on traditional primary product exports. Among the factors cited as the underlying reasons for this proposed shift are' (1) the purported deterioration in the terms of trade for primary commodities; (2) the substitution of synthetics for many of these items (plastics for metals, artificial for natural fibers, chemical sweeteners for sugar); (3) the instability of primary product pxices in international markets; (4) the increased employment opportunities associated with the production and export of manufactures; and (5) the realization of economy-wide linkages and *learning effects' resulting from the processing (manufacturing) function (see Helleiner and Welwood, 1978 or Roemer, 1979). One method suggested for increasing the proportion of developing countries' trade in fabricated goods is to increase the processing of natural resource-based products now exported in primary form. However, a factor often cited as worling against efforts to increase domestic processing is the structure of tariffs and other trade barriers in major import markets. Specifically, zero or low tariffs are generally applied to industrial countries' imports of primary (unprocessed) commodities with the duties increasing, or 'escalating', as the product experiences increased fabrication. Tariff escalation has been acknowledged to produce a tmde bias against processed goods due to the higher import duties imposed on these items.' 'For example, Balassa (1968, p. 195) states that 'increases in nominal and effective duties from lower to higher stages of transformation point to the existence of discrimination against the processed export products of developing countries.- A similar position concerning the influence of escalating t- Yfs has been taken by Johnson (1965). Representative studies which document the existence of tariff escalation -a developed countries include UNCTAD (1968, 1980) and Yeats (1979). Yeats (1984) argues that a trade bias against processed commodities may occur even when there is no escalation in tariffs because of generally higher import demand elasticities for processed as oppo99d to primary commodities. 2 The importance that developing countries attach to trade barrier escalation is reflected in the extensive policy debates on this subject that have occurred in major intemational forums. For example, developing countries wore instrumental in having a plank inserted in the 1982 GATT Ministerial Declaration (p. 16) stating that "prompt attentior should be given to the problem of escalation of tariffs on products with a view to effective action toward the eliminat.on or reduction of such escalation where it inhibits intemational trade, taking into account the concems relating to exports of developing countries. ' The Punta del Este Declaration also stated that 'negotiations shall aim to achieve the fullest liberalization of trade in natural resource-based products, including those in processed and semi-processed forms. The negotiations shall aim to reduce or eliminate tariff and nontariff masures, including tariff escalation.' UNCIAD (1979), Commonwealth Secretariat (1982), World Bank (1981 and 1987) have also viewed tariff escalation as a major problem for developing countries. While trade barrier escalation has been an important point of contention in the Uruguay Round and previous multilateral trade negotiations (MTNs), the topic's relevance to intra-Asian has not been clearly established. Yet a number of Asian countries like Astralia, China, Malaysia, Philippines and Tailand have a major interest in promoting further processing of natural resource products that are now often exported in raw or semi-processed form (see Table 4). However, almost aU of the empirical studies that have documented the existen of trade barrier escalation, and the resulting import bias against processed goods, have focused an Japanese, North American and European markets and it has not been established that developing countries' trade barriers follow a similar pattern.' Moreover, it has not yet been determined that developing countries' (purported) naturnl advanta in primary commodity production generally establishes cost diffeentdals of a sort tlrt salating trade barrier protection is not required for domestic processing industries. The purpose of this study is to assess the priority that should be assigned to trade barrier escalation in any multilateral libealization effort involving Asian countries. Employing a commodity processing chain classification scheme developed by the World Bank (see Box I for an example of a commodity processing chain), the structure of individual South and East Asian countries' imports is analyzed to determine if a significant trade 2One exception is a study by Laird and Yeats (1987) which showed that tariffs in 23 developing countries or country groups often incorporate a high degree of escalation. However, the major focus of this study was on countries outside the Asian region so it is not directly relevant to an Asian Round. 3 bias exists against semi-fabricated and processed commodities. Next, using detailed information on tr&de barriers compiled by UNCTAD and the World Bank an attempt is made to determine if Asian countries' tariffs and nontariff barriers (NTBs) escalate in the same manner, and to the same degree, as developed countries' trade restnictions. The analysis also attempts to identify specific commodity processing chains where liberalization could make an important contribution to Asian intra-trade. The study concludes with an overall assessment of the priorities that should be assigned to trade barrier escalation issues in Asia and also provides several specific suggestions for a regional multilaterml liberalization initiative. II. The Data and Methodology In this study a World Bank classification scheme was used to identify different levels of fabrication for 48 commodities exported in primary and processed form by Asian developed and developing countries (see Appendix I for full details on the components of each stage identified in terms of SITC products). At a minmnum, the scheme distinguishes between a primary and processed stage product (i.e., the primary stage of the coffee chain consists of green and roasted coffee beans (SITC 071.1), while the processed stage consists of coffee extracts (SITC 071.2)). In other instances, a semi-fabricated stage or stages are identified (i.e., the cocoa chain consists of cocoa beans (primary stage), cocoa powder and butter (intermediate stages), with chocolate being the final stage item). Table 4 contains a ist of commodities covered by this classification scheme, and identifies major Asian country producers and exporters of these products.3 Statistics on Asian countries' 1990 imports of each processing chain's stages were drawn from United Nations Series D trade tapes, as were similar data for several earlier years. In a few cases, trade data for individu2i commodity processing chains were aggregated into broad product groups (e.g., foodstuffs; ores, minerals and nonferrous metals; or energy products) to focus on broad trends in primary and processed products' import 3"Leakages me y occur as one goes up the processing chains since some products may be employed as inputs for production processes which are not a formal part of the chain. For example, cotton fiber may be used for the manufacture of rubber tires and not appear in the textile stages of the cotton chain. Such leakages may cause the magnitude of the shift to processed commodity imports to be understated and may also affect the accurv'y of estimates of the es-alation in tariffs and nontariff barriers. If trade barriers are lower for products where leauges occur than for goods included in the formal definition of the processing chain, the extent and magnitude of escalation will be overstated. 4 shares. Due to differences in factors like the level of development, the import structures (and trade barriers) of countries in South Asia, non-OECD) East Asia, and OECD Asia are analyzed separately. Table 1 identifies the major exporters in each group while the World Bank (1992, Table Al, p. 40) provides a full listing. Statistics on Asien countries' trade barriers were drwn from two sources. Tariff data were compiled directly from GAIT documents and then compute.ized and stored in the World Bank-UNCTAD SMART (Software for Market Analysis and Restrictions to Trarle) data base. (See UNCTAD-World Bank, 1989 for a description of the SMART system). Information on Asian countries' nontariiff barriers was drawn from UNCTAD records and also incorporated in the SMART system. Since the tariff and NTB dati were recorded at the tariff line level, available concordances were used to aggregate these data to Standard International Trade Classification (SITC) groups. This procadure allowed the UN trade data to be matched directly with Asian countries' tariff and NTB infornation. Im. The Coimmodity Structure of Asian Trade Three key points should be considered in an assessment of escalation issues: (1) whether Asian countries' imports reflect a major bias against semi-finished and processed products; (2) whether a similar bias is reflected in Asian commodity exports; and (3) whether Asian trade barriers escalate. If these conditions are found to exist, the subject of escalation should be addressed in Asian trade negotiations.' 'An important conceptual atl ampiical issue is how to deternine if a bias exists in Asian trade and, if it does, what its magnitude is. In this stusy, the _ mnmodity import structure of the European Community (EC) is sometimes used as a "standard' for comparison. Two po - - should be noted with regard to this choice. First, Balassa (1968), Yeats (1984) and others argue " at a signifiu.. ias against processed commodities exists in EC imports and have assessed the influence of contributing factors. Here, any Asian import bias is measured relative to that of the EC. Second, separate tests conducted by the authors show #hat the overall commodity import structure of the United States is quite similar to that of the EC although there are some differences for several individual commodities (see Box 2 for details). Thus, the conclusions of this study would not differ significantly if the United Statee rather than the EC were employed as the standard for comparison. 5 iox 1 Characteristics of SC-Based C.i'uuodity Processing Chains Although the approach has several recognized imperfections, a number of empirical studies have utilized the Standard International Trade Classification (SITC) system to construct "commodity processing chains" (Balassa 1968, Yeats 1979, Laird and Yeats 1987). Thesw chains trace individual commodities like cocoa and coffee beans, or cotton, jute and iron ore, through successive stages with each experiencing a greater degree of processing then the former. After detailed analysis of the SITC system, the World Bank was able to construct processing chains for 48 different commodities (see Appendix 1 for details). In 1990, the individual components of these chains accounted for approximately 85 percent of all developing countries' exports. The major advantage of an SITC-based processing chain framework for analysis is that it provides a common basis for merging trade and trade barrier information. That is, the processing chains allow one to tabulate a country's imports or exports of each stage in a given chain and match this information directly with statistics on tariffs and nontariff barriers. The following example shows the three stages of the lead processing chais, Japan's 1990 imports at each stage, as well as Japan's average (unweighted) tariff on each product. Note how the import duties increase or *escalate" as one move: from lead ore to wrought lead alloys. 1990 Japanese Imports Processing Stage (SITC) Value (Smill.) Share (%) Tariff (%) LEAR Lead Ore (283.4) 120.3 57.1 0.0 Unworked Lead idloy (685.1) 87.9 41.7 3.2 Wrought Lead Alloys (685.2) 2.5 1.2 558 All Stages 210.7 100.0 Two specific points should be noted concerning these SITC-based processing chains. First, the level of detail changes from chain to chain due to the nature of the SITC system. That is, sevedal chains like cocoa and wood pulp contain three and four stages wbile other commodities like coffee or copra only include an unprocessed and processed stage. Second, there is a problem of "leakages" in some chains, L.e., some commodities may be production inputs for items not included in the chain's components and lost from the analysis (see footnote 3 for a specific example). It is acknowledgeJ that the problem of leakages may produce biases in SITC-basod analyses of trade barrier escalation. 6 Table 1 presents summary information on 15 major. Asian countries' commodity imports in 1990. The reported data are aggregates of all 48 individual commodity processing chains (see Appendix 1) that have been classified under four broad headings (agricultural materials, foods and feeds, ores and metals, and energy products). The first five columns of the table show -acih country's imports of a" pnmary and processeJ stages of the chains while the next fivo show the share of primary (unprocessed) stage products in total imports. Finally, the individual country results are aggregated into an overall average for Asia (see the meme~ item) and thbs information is compared with similar statistics on EC imports.' Overall, Table 1 shows the Asian countries' imports are considerably more biased against processed commodities than those of the EC. This is surprsing given that the EC has often been criticized in UNCTAD and GATT for its (purported) bias against processed commodity imports. Approximate'4 36 percent (by value) of EC commodity imports are composed of primary (unprocessed) stage commodities while the corresponding share in all Asian countries is 16 percentage points higher (52.3 porcent). Relatively higher Asian import concentration in unprocessed commodities occurs for all four product groups, but is especially pronounced in ores and metals and agricultural materials where the Asian share is, respectively, two and threa times that of the EC. A key related point is that the overall import bias reflected in Table 1 also shows up in Asian intra-tade. Asian countries now export a considerably greater share of processed commodities to non-regional markets than to Asian markets (see Box 2). Table 1 also reveals considerable variation in individual country results. Japan records the second highest (behind India) overall degree of prmary commodity import concentration (62.7 percent of Japan's total MThesw overall averages can be affected by an Asian country's natural resource endowments. If, for example, a country has abundant natural resources it might tend to have relatively lower processed good imports, ceter* paribus, than countries not so well endowed if resource availability stimulates domestic processing. International transport costs could work toward this end since many bulky low value prmary commodities often have relatively high nominal freight costs which may decline significantly with further processing, although this is not always the case (see Yeats 1977). For some commodities like metal ores, stowage factors decline sharply with processing which makes the fabricated product easier to transport. For these reasons there may be a tendency for processing activity to be located close to available supplies of some natural resoures. Of course, escalating rade barriers in major international markets would be an offsetting factor. There are some products, however, like wood manufactures th4A become more fragile or subject to pilferage after processing than the primay commodity input (wood). For such items transport costs may have an insignificant effect on the location of processing activity. 7 Table 1. 'Mb Relative Imponawce of Asian Couniee Imports of Primary *nd Ptocuead Commodliies in 1990 Vaha of AlR Frimary Stage and Ftcesued Commotes (Smillion)' Shar of Pinmay Stap Podus in ToalT b4mpos (%) Foods and Agnicltural OCa and Energy AUl Foods and Agricuhural Ors and Energy AUl ImPorter (Year) Feeds Material Mtas4 Products Connodities Feeds Matrias Metals PIoducts Conuoditics' South Asi. 1,853 1,736 2,389 4,587 10,565 74.3 40.0 23.4 83.6 63.1 (64.6) Ini (90) 479 1,135 1,828 3,342 6,785 86.8 30.3 26.0 99.0 70.8 (75.4) Paldsan (B9) 999 319 326 1,144 2,787 72.6 32.0 21.S 3i.0 47.4 (44-7) Sri Lanka 490) 375 282 23S 101 m 62.9 7.7 5.9 100.0 54.7 (S63) Non-OECD East Asi 18,139 26,840 28,492 32,669 106,139 64.7 36.4 17.6 63.4 43.1 (52.1) China (90) 3,244 7,098 3,261 9S2 14,SS4 83.6 25.8 61.5 44.5 38.2 (54.0) Hong Kong (90) S,000 1,479 1,309 2X6 17,014 43.9 7.8 1.7 0.0 16.9 (15.2) lodonesia(90) 745 1,270 1,381 1,821 5,218 91.5 47.8 7.3 64.4 49.2(61.9) Rep. of Korea (90) 2,352 7,342 4,146 8,761 23,301 69.8 53.1 40.2 72.8 57.2 (63.4) Malaysia (90) 1,207 1,102 1,S66 1,227 5,101 71.9 21.3 14.5 13.0 29.2 (32.8) Philippines (88) 462 392 536 1,051 2,441 73.1 24.2 22.5 91.5 62.1 (62.2) Singapore (90) 2,464 2,421 2,175 9,486 1656 48.5 25.6 2.5 73.0 53.2 (46.2) Taiwan, China (90) 1,632 3,595 3,926 4,310 13,463 69.1 40. S.? 73 8 453 (56.8) Tbailand(90) 1,033 2,141 2,492 2,835 8,501 94.1 43.2 2.8 53.7 41.0(7.3) OECD Asia 26,158 28,860 20,92P 33,85 109,232 63.5 30.3 37.6 97.8 60.3 (64.8) Aurlia (90) 1,057 2,946 814 1,748 6,566 47.7 4.0 21.0 59.4 27.9 (38.5) Japan (90) 24,733 25,389 19,S55 30,99S 100,672 64.3 33.8 39." 99.9 62.7(66.7) New Zealand (90) 367 525 S60 542 1,994 53.1 7.4 11.3 100.0 42.1 (S4.0) Menmo hem EuropeanCommunity (90) 90,671 104,053 61,534 81,045 344,303 52 7 11.1 '2.2 70.6 36.3 (44.5) All Above Asan Countries 46,149 S7.436 51,810 70,S41 225,936 64.4 36.3 25.7 80.0 52.3 (58.7) 3ee Appendix I for the primary and proceed stage products classified in eah group. Figuresinparuthesesshowaveragesbaedenweighutleing theindividualcommodity group'apone in worldtrad. Ao oher gursare basedonthe Asiancounty'sowntradeweighu. Soure: Trade data compiled from United Nations Statiw;cl Offie recors. 8 Box 2 Do Asian Processed Commodities Fare Bettwr I Non-Asiar Markets? Tables 1 and 2 show a major Asian import bias exiss against processed somodities and, as a result, many Asian countries export a relatively high share of their domestically produced commodities in primary form (see Tabl,.s 3 and 4). This observation raises the question of whether the structur-I bias against Asian processed commodities exists in non-Asian markets and, if so, whether it is greater or smaller than in Asia. The followng tabulations comparm the total value of othar countries' 1990 imports of primary and pwcessed commodities from Asia and afro indicate the share of this exchange which consists of unprocessed goods. To assist in comparisons, the tabulations provide similar statistics on Asian intra-trade, and on trade with selected OECD markets. Food & Ores & Agricultural Energy Al Importer Feeds Metls Materials Products Items (Value of all impor4 from Asia - USS million) ALL-ASIA 20,692 21,297 33,575 20,882 96j446 Japan 10,924 7,702 9,534 7,487 35,654 Korea 825 2,372 2,731 2,199 8,128 NON-ASIA OECD 13,742 6,778 26,375 2,929 49,814 U.S.A. 5,826 3,369 11,707 2,637 23,540 Canada 783 395 1,045 18 2,242 Oermany 1,563 1,026 2,979 78 5,646 U.K. 1,313 561 2,061 31 3,996 Sweden 152 60 338 - 550 (Share of prinmay stage in all imports from Asia - percent) ALL ASIA 63.2 25.9 28.5 64.5 43.2 Japan 64.4 53.5 42.8 Q9.9 63.7 Kvrea 64.7 26.8 52.7 57.9 47.8 NON-ASIA OECD 52.1 13.5 17.3 75.0 31.2 U.S.A. 43.9 4.1 9.1 76.2 24.5 Canada 47.5 12.9 8.6 98.3 23.7 Germany 56.5 51.1 21.6 77.4 37.4 U.K 47.6 31.1 14.5 - 27.7 Sweden 40.7 30.9 3.7 - 16.9 The above comparisons show Asian processed commodity exports fare considerably better outside the region than in intra-Asian trade. Specifically, 43 percent of all Asian intra-trade consists of primary stage products, a share which is 12 percentage points higher than exports to non-Asian OECD markets. Why is the share of processed commodities in intra-Asian trade so low? The above tabulations show that Japan's import performance (and to a lesser extent the performance of Korea) has a strong influence on the overall result. Both countries ar major commodity importers and both have a clear bias against processed goods (i.e., 64 percent of Japan's total imports consists of primary stage products). This observation attaches special importance to the level and sttucture of Japanese and Korean trade barriers. 9 Table 2. Trends in the Relative Importance of Asian Imports of Primary and Processed Commodities: 1970 to 1990 Commodity Groug Importing Region Year Foods and Agricultural Ores and Energy All Feeds Materials Metals Products Commodities (value of total imports of primary and processed commodities - USS iillion) South Asia 1990 1,853 1,736 2,389 4,587 10,56S 1980 1,453 S61 1,522 3,251 6,787 1970 748 281 307 39 1,374 Non-OECD East Asia 1990 18,139 26,840 28,492 32,669 106,139 1980 7,977 10,532 7,359 26,678 52,546 1970 1,423 1,;58 663 893 4,347 OECD Asia 1990 26,158 28,860 20,928 33,285 109,232 1980 12,024 19,267 13,080 61,378 105,7S1 1970 2,231 4,675 3,311 2,996 13,213 All Above Asian 1990 46,149 57,436 51,810 70,442 225,837 Countries 1980 21,454 30,360 21,961 91,307 165,082 1970 4,402 6,324 4,281 3,928 18,935 (share of primary stage in all imports of the group - percent) South Asia 1990 74.3 40.0 23.4 83.6 63.1 1980 33.1 25.3 14.7 13.6 19.0 1970 62.6 62.3 16.2 0.0 50.5 Non-OECD East Asia 1990 64.7 36.4 17.6 63.4 43.1 1980 61.6 52.3 16.7 79.0 62.3 1970 63.1 51.9 15.3 57.4 51.1 OECD Asia 1990 63.5 30.3 37.6 97.8 60.3 1980 76.9 47.9 56.0 86.4 74.5 1970 85.0 56.1 70.6 74.0 68.7 All Above Asian 1990 64.4 36.3 25.7 80.0 52.3 Countries 1980 68.2 49.0 40.0 81.6 68.3 1970 74.1 55.5 58.1 69.5 63.3 Memo Im European Economic 1990 52.7 11.1 12.2 70.6 36.3 Community 1980 57.9 16.3 16,3 76.6 51.1 1970 60.6 25.3 20.7 82.6 46.1 Note: See Table 1 for the countries included in each regional group. Appendix 1 identifies the individual products classified as primay stage items in each processing chain. Source is as for Table 1. 10 imports are primary stage items), whiie the Korean average (57.2 percent) is also well above that of other Asian countries. Neither Japan nor Korea figures prominently among important Asian commodity producers (see Table 4). Their combined imports of primary and processed commodities are almost 55 percent of the Asiae total. These facts highlight the importan .e of tariff escalation issues in any multilateral Asian trade liberalization effort. Since Table 1 shows that the current structure of Asian imports is heavily biased against processed commodities (whether these shipments originate in Asia or elsewhere), an important related point is how this structure is evolving over time. Table 2 addresses this point by tabulating the value of Asian imports of primary and processed commodity imports at ten year intervals over the last two decades and also by showing the share of unprocessed (stage 1) commodities in this trade.6 The import performance of the European Community for these same goods over the same period is also reported (see memo item). Between 1970 and 1990 the share of Asian imports of all unprocessed commodities declined by approximately 11 percentage points (from 63 to 52 percent), which is about the sanm as the decline registered by the EC. In other words, the relative Asian trade bias against processed commodity imports has not narrowed over this 20-year period. This finding accents the -eed for policy action to identify and remove existing constraints to Asian intra-trade in processed commodities. Aside from the overall regional trend, several different patterns are reflected within the regional sub-groups. South Asia's import shares for unprocessed agricultural raw materials declined sharply betwoee 1970 and 1980, but this was primarily due to special factors not directly connected with a shift to processed products. For example, India all but discontinued its (previously major) imports of raw cotton while its imports of cotton thread and fabrics remained fairly constant in dollu terms- this produced the appearance of a shift toward fabricated cotton goods. Historically, India was a major market for Egyptian long-stem cotton - dia's imports averaged 25,000 tons per year in the early 1970s - but, after developing domestic production of a suitable cotton substitute for spinning fine yams, its imports of Egyptian cotton declined to an average of less than 3,000 tons per 'One important data problem to be noted is that it was not possible to hold the composition of importing Asian countries constant over the 1970-90 period. The Peoples Republic of China did not report imports or exports to the United Nations prior to 1984 so the group of Asian countries included in the 1970 and 1980 statistics is somewhat different from that for 1990. Tables 1 and 2 rely on trade data reported by individual Asian countries (there are gaps in some country's reporting) while Tables 3 and 4 show imports reported by their trading partners. For this reason it was possible to report a somewhat larger selection of smaller Asian countries like Bhutan, Nepal and the Maldives in the latter tabulations. 11 year. India also imported major quantities of wheat meal and flour on a temporary basis in the late 1970s and early 1980s for famine relief. These imports were subsequently halted and this caused the unprocessed foods import share to decline by about 40 percent over the period 1970-80. The changes in crude and refined petroleum prices have also affected the inter-temporal comparisons. The fact that petroleum is largely traded in (crude) unprocessed form, coupled with relatively high prices in 1980 and 1990, greatly increased the weight of this commodity chain relative to all others. As a result, the progress that was made in shifting trade in some chains toward semi-finished and processed goods is masked in the overall averages. In Table 3 the focus of analysis shifts from the import performance of Asian markets to Asia's export experience in prmary and processed commodities. The table addresses thuee questions: (1) how important are primary stage commodities in total exports (including all items whether or not they wero classified in a processing chain); (2) which Asian countries are the largest (absolute) exporters of prianry commod ities, and would therefore potentially experience the largest benefits from a reduction in barriers facing procossed goods; and (3) how has the concentration in primary commodity exports changed between 1970 and 1990. Table 3 reinforces the importance of addressing tariff escalation issues. Primary stage commodities account for at least one fifth of total exports for one half of the 30 listed Asian exporters, while these items comprise at least 15 percent of total exports for 22 of the 30 countries. Australia is the single largest exporter of pdimary stage commodities with $9.6 billion in trade, while 12 of the 30 Asian countries have total primazy commodity exports of $1 billion or more.' A second point to emerge from Table 3 is that primary stage commodities are particularly important for several smaller countries like Brunei, Fiji, Guam and Tonga where export shares range between 43 and 59 percent of total exports. (The share of these exports reaches a high of 73 percent for Papua New Guinea). Furthermore, the share of primary commodities in total exports has actually risen for several of these countries between 1970 and 1990 (e.g., Tonga, Vanuatu, the Maldives). These countries have as much (or even more) stake in negotiating trade barrier escalation issues as do Australia, China, Indonesia or New Zealand whose total absolute 7If China's exports are combined with those of territories on which it has claims, i.e., Taiwan (China) and Hong Kong, its total primary stage commodity exports would be roughly $11.3 billion, or $1.7 billion more than Australia's exports. 12 Table 3. The Relative IMportnce of Primay (Uaproceused) Stage Commodities in the Total Export of 30 Asian Countries: 1970, 1980and 1990 Country's Shame In Asian Nominal Primary Stage Conmmodides' Exposb of All Primry Stage Value of Aii Primary Staep Share of Totai Export (%) Commodities (%) Comnodity Exports (Smiilion) Asian Expoter 1970 1980 1990 1970 1980 1990 1970 1980 1990 Ausalia 28.6 27.4 25.5 23.8 12.6 18.5 1,397 5,784 9,608 China 23.7 25.4 10.5 7.2 10.1 16.7 420 4,628 8,670 Indonesia 61.5 62.6 33.5 14.1 31.5 16.7 829 14,423 8,666 Malaysia 44.5 39.3 18.6 7.6 13.2 11.7 446 6,033 6,101 TMailand 35.6 31.0 18.5 4.4 4.4 7.8 258 2,006 4,050 India 31.9 20.9 14.5 9.9 3.6 4.4 579 1,664 2,291 Taiwan, China 10.5 4.8 3.1 3.0 2.1 4.2 176 947 2,184 New Zeaand 14.7 17.0 22.7 3.4 2.1 4.0 200 966 2,090 Philippines 29.7 29.7 16.1 6.2 4.3 2.9 366 1,969 1,508 Rep. of Kora 14.3 4.2 2.2 2.0 1.4 2.4 115 6J6 1,254 Btunei 93.0 62.9 43.7 1.8 5.7 2.0 106 2,586 1,035 Japan 1.6 0.6 0.3 4.4 1.7 1.9 260 785 978 Papua, New GuInea 68.9 83.1 r,.2 0.9 1.9 1.3 51 881 657 Singapore 17.1 4.5 1.6 1.9 1.4 1.2 112 620 626 Pldstan 20.2 18.6 12.6 2.0 0.9 1.1 120 423 595 Sri Lanka 69.9 45.4 21.0 3.6 1.0 0.7 209 469 381 Hong Kong 1.0 1.1 0.7 0.4 0.4 0.6 25 180 322 angladeb n 23.4 17.4 na 0.4 0.6 as 199 295 F3ji 71.4 56.4 50.5 0.8 0.4 0.3 45 166 168 Korea, Democratic Rep. 243 6.2 17.3 0.3 0.1 0.3 17 S0 131 Myanmr 12.8 33.1 19.7 0.3 0.3 0.2 15 156 98 Guam 333 31.2 46.8 - - 0.1 1 1 29 Maldives 33.3 36.4 36.8 - - 0.1 2 4 28 Democratic Kampucbet 43.6 28.6 48.6 0.3 - - 17 2 18 vanutu 90.9 56.6 69.2 0.2 - - 10 17 18 Solomon Ilans 37.5 47.4 21.0 0.1 - - 3 37 17 Mongolia 3.5 29.8 23.3 - - - - 2 16 Tonga 97.3 55.6 58.8 - - - 1 5 8 Nepal 27.1 18.4 2.0 0.1 - - 7 14 5 Kinbati 99.5 66.6 66.6 0.1 - - 6 4 2 Source: Compiled from United Nation's COMTRADE ecords using partner country imports. 13 value of commodity exports is far greater, In short, Table 3 shows conclusivel: that the potential interest in further Asian processing of primary commodities is spread among a large number of countries. Table 4 profiles the primary commodities in which each Asian country has an important interest. These tabulations identify the largest exporters, sbow their share of all shipments of each commodity, and indicate the value of exports originating in South Asia, OECD Asia and non-OECD East Asia. The intention here is to discover how dispersed Asian interests are across regions, commodities and countries.' We want to determine whether major exporters have a stake in escalation issues for one or two commodities only, or whether their interests are likely to be spread over a wider range of items. The general impression from Table 4 is that Asian exporters' interests are dispersed over a fairly large number of commodities. This has positive implications for any multilateral trade negotiations in the Asian region since it creates greater opportunities for trade-offs. Australia and China, for example, rank among the principal suppliers of at least 16 of the 42 commodity chains listed in the table (several relatively unimportant chains like hom and bone have not been included) with China being a factor in 25 chai. India, Indonesia, Papua New Guinea, the Philippines, Malaysia, New Zealand, Taiwan (China) and Thailand are among the largest suppliers of five or more commodities. A further point to note is that Japan, which would play a major role in the negotiations given its prominence as the largest import market for commodities (see Table 1), does not figue as an important primary stage exporter for any of the commodity chains listed in Table 4. This observation accents the need for analysis of the structure of Japan's trade barriers. 'Overall, 70 percent or $38.3 billion of the primary commodity exports originate in East Asian non-OECD countries, 23 percent ($12.9 billion) come from OECD Asia, and 7 percent ($3.9 billion) originate in South Asia. However, these comparisons understate somewhat the true importance of commodities for the latter as these shipments account for roughly 20 percent of all South Asia's exports as opposed to 14 percent of the total exports of non-OECD East Asia. '4 Table 4. Major Asian Exporters of Primary (UProcesseO Stage Commodities in 1990 1990 Asian Eors of the Commodity (Smill.) South Non-OECD OECD All Primary Stage Commoditv (SYIC) Asia East Asia Asia Asia Major Asian Exponers (Share of Total Ain Exponls - in percent) 1. Foodstuffs Shel frsh 031.3) 792 4,168 681 5,641 China (21), Thailnd (19), Indonesia (12) Fish, frsh or child (031. 1) 101 2.907 620 3,628 Tiwan, Cbina (26), Koe (19), China (10) Vegetablks, fresh (054) 85 2,218 352 2,655 Thiland (34), China (23), Taiwan, China (12) Frun and nuts, fr^esh (051) 280 1,135 856 2,271 NwZland (29), 1hipmnes (21), Chit (13) Raw bedt and cane auger (061.1) 7 964 757 1,728 Auslia (44), Th,ilanl P0 Philpines (7) Unmilledwheat (041) - - 1,159 1,159 Ausalia (99) Tea (074. 1) 467 S06 4 977 China (35), Sri Lank (25), India (21) Coffee, green or asted (071.1) 76 703 2 781 Indonesia (52), Papus New Guinea (16), lTailand (15) Cocoa buans (072.1) - 356 - 356 Malaysia (61), Inoa (2), Papua New Guinea (11) Uve wine (001.3) - 288 - 288 China (74), Malays (26) Grounduuts (221.1) 5 279 - 285 China (85), Taiwan, China (10) Soyabeam C221.4) 1 207 1 210 China(93), Taivan, China (2) Lve poultry (001.4) - 171 3 174 China (55), Ma a (41) Live cows and sheen (001.1, 001.2) 1 29 82 112 Ausrali (63), Cina (24), New Zealand (10) Rice inhusk (042.1) 36 32 12 s0 Thailand (39) India (38), Austalia (1) Copra (221.2) 4 64 - 67 Ph7ilippines (3), Papua New GuuQ7 Paln nuts (21.3) - 4 - 4 Papu- New Guinea (52), Malay (45) Lineed (221.5) 1 - - 2 In& (18), New Zeand (15) B. Agricultural Materials Rubber(231.1, 231.2) 69 3,811 472 4,353 Malysia (37).Thail (22), Indonesia(20) Rough log (242.21, 242.31) 167 2,768 282 3,217 Malaysa (77), New Zealand (S), PFpua New Guinea (4) Couon(M63.1,263.4) 81W 252 525 1,593 AutAia(33),Pakistan(28),India(23) Wool (262.2, 262.6, 262.8) 13 141 1,172 1,326 Austalia (58), New Zelmd (30, Malaysia (3) Hides and kins a(2l1 Ies211.8, 211.9) 3 147 736 885 Aualia (49), New Zealand (34), China (14) Raw silk (261.3) - 488 12 50 Chia (83), Taiwan, China (8) Feathers (291.96) 1 366 2 370 Chin (46), Taiwan, China (34), Hong Kong (9) lute(64) 94 27 - 121 Banltab (76),1China(1) Hom and whalebone (291.12) 9 17 IS 41 New Zland (37), China (17), India (17) Flax, hebanp anie a6s.1 to 265.3) - 18 1 19 China (74), Taiwn, China (16) UnwokedLmm hair (291.91) 2 8 - 10 China (60), India (20 XI. Ores and Minerals Ironm e81.3) 723 241 2,510 3,474 Atutrlia (72), ndi (22), Phipies (6) Copper or (83.11, 283.12) 7 1,208 192 1,408 Papua New Guinea (32), Idonesia (27), hilippines (21) Zinc Ore (83.5) 1 25 49S 521 Ausralia (95), China (3) Bauxite or (283.3) - 139 214 352 Ausria (59), China a(8) Mangane ore (283.7) 15 6 224 246 Ausalia (91), India (6) Sand, excluding meta beaing (73.3) 1 40 el 122 Austalia (63), Chia (15) Tin oe a83.6) 3 59 44 105 China (SO), Ausralia (42), Maysia (2) Lead ore (283.4) 1 IS 65 81 Anamra (80), Thailad (IS), Koes (2) Tungtn ore 283.92) 1 33 5 39 China (77), Aus;tia (13) Natural phosphate (271.3) - 12 3 IS China (67), Australia (0. Singaooe (7) Silver ore (285.01) - 2 - 2 Indonesia (99) Cade abesto (276.4) - I - 2 Sinppore (50), Cbina (30) IV. Enemyhw Cmde petroleum (331.01) 74 14,469 1,371 15,914 hIdonesia, (36), Ca (24), Malaysia (23) Source: Compiled fiom United Naio' COM1RADE record using t% -ad part cury iaoma. 15 Box 3 Can a Trade Bias Against Processed Goods Still Occur if there is no Escalation of Trade Barriers? For trade negotiations, it is important to distinguish between two points: trade barrier escalation which refers to tariffs and NTBs rising with fabrication, and the influence of these barriers on the structure of trade. To account properly for the latter, one must analyze changing conditions of demand at different levels of processing. Since empirical studies show import demand elasticities normally increase with fabrication, constant tariffs will have relatively larger trade effects on fabricated commnodities than on unprocessed commcdities. This point can be clarified through the use of the example cited below. Here, it is assumed that the leather processing chain is composed of three distinct stages (hides, leather and leather manufactures), and import demand elasticities range from 0.6 for hides to over 2.0 for leather manufactures in developed countries. For illustration, it is assumed that the importing country ap2lies a constant 10 percent tariff and imports $20 million in each processing stage. There is no tariff escalation, yet the tariff has more of a retarding effect on leather manufactures due to the more 'sensitive demand for these products. Specifically, reducing the tariff for hides to 5 percent would increase imports by $558,000. A similar cut applied to leather manufactures would increase imports by more than three times this amount. Thus, in assessing the influence of trade barriers, consideration must be given to underlying demad conditions to draw meaningful conclusions about their influence on the trade structure. Stated differently, the escalation of trade barriers is normally a sufficient condition to conclude that an import bias against processed goods exists, but it is not a necessary condition. Projected Change Processing Nominal imports Import demand with a 50% Stage Tariff (S mill.) elasticity tariff cut ($000) Hides and skins 10 20 -0.62 558 Leather 10 20 -1.28 1,152 Leather mfg. 10 20 -2.11 1,899 Note: The projected import change is based on the use of a partial equilibrium trade model in which the estimated change in imports is derived from a multiplicative function involving the initial level of imports, the import demand elasticity, and the clange in the landed price of the good due to the tariff reduction. See Stern (1976) for a discussion of this type of analysis. 16 IV. The Structure of Asian Trade Barriers The analysis to this point has demonstrated that Asian commodity imports reflect a strong bias against processed goods (Table 1), and that the bias against Asian exports of processed commodities is greater in intra-Asian trade than in non-regional markets (Box 2). Furthermore, between 1970 and 1990, Asian markets' import bias showed no evidence of narreiwing relative to other markets (Table 2). These facts raise the question of whether, and to what extent, this sub-par performance can be accounted for by escalation in Asian trade barriers.9 Table 5 draws on GATT tariff data for 10 Asian countnies in order to provide relevant information." The table shows the average import duty each country applies to the primary stage item in the 48 commodity chains, as well as the tariff on the final stage product. Tariff averages have been computed for all 48 commodity chains and their differences are used as an overall measure of escalation for each market. It is clear from Table 5 that ther are major differences in Asian tariff levels, and that tariff escalation occurs in most processing chains. Concerning the first point, GATT records indicate that Hong Kong has no tariffs on any primary or processed commodity (for this reason Hong Kong is not included in Table 5), while Singapore has minor duties of under 4 percent on seven processed commodities (refined petroleum is an exception with a 10 percent tariff). In contrast, tariffs of over 30 percent occur in Australia, Indonesia, Japan, Korea and 'It mw: te acknowledged that studies have identified numerous other problems besides external trade barriers that developing countries face in trying to increase processing of domestically produced commodities. These include: inappropriate economic policies (import substitution) pursued by the country itself; insufficient access to intemational capital or technical markets; tansnational corporation policies; monopoly pricing practices of liner conferences that may increase freight costs; and anticompetitive practices of established firms in major OECD markets. For a survey of these problems see UNCTAD (1979). It should also be acknowledged that some countries' comparative advantage may be in exporting unprocessed commodities and not in processing these commodities. 'The underlying tariff data were compiled by GATT and then recorded in the UNCTAD-World Bank SMART system. No data on China's tariffs were collected since this country currently is not a member of the General Agreement. Due to the magnitude of the effort required to compile matched trade and tariff statistics for a single country - more than 12,500 tariff line level products may be involved - GATIT did not attempt to compile statistics for several of the smaller Asian countries that are GATT members. UNCTAD's Inventory of Trade Control Measures was the source of information used in this study on nontariff barriers - see Table 6. Tabic S. Compamof Asin Cwmui' Tuiffi an Phimy Fmil Stea of 4B Cowaduty Pmoceeh Chaia. grRAIJT A R40033IN A * APAN KOOEA MALAYSA NEW ZEALAND P SNAPORE 7TAAND 9 ..a, 0b P_ ay Fb3 Prdmy Fail PHkm) Fbi h y ey F kpI P.ksy Fbi Phimy FbI hi=w Fbi y Fdi I McMe d l o e. bm lmi.(3) 3.3 S.9 6. 43.3 OA 17.3 12. 27.1 7 A S.3 0o 144 l0.0 JOB 0.0D 24D D DD 2 swis 0) 0D0 7.9 1.5 40.9 OA 138 2eD 31 0.0 9 0.0 14.1 10D0 sD. OA 0.0 2DD Go 3 r.k.y 4) O 8.3 135 1 0.0 1eD 13.3 31.3 SD 9D OA U.2 4*O JOB 0D OA 20D 6D 4 Fbek83ham 0* *ID 29.1 46.2 5.5 OA 239 3.4 0.0 36.1 12 5g 2DD 42.0 0o OD 6e D .0 5 S33&b b.b d6bdt, 0A OA) 30D 35.0 79 67 23* 3*10 27 44.3 OD 42L sDD 2* OD* D 6.0 DD 6 WeaD OA 8.4 0. 33. 230 2 7.2 s 30D 1.3 33.7 O. 2se I 0D 4.0 OD 3.7 OA 0D 7 RIm 2.0 2* 0.0 0O D lsD 5* 5* SJ 0* 0. O* e0* o 0 93.0 0* 0* 0.0 0* 8 Fo& 20 112 3D2 s e 3sD2 4* 5eD fGA 0* Z5J 0* 19J sDD 43J 0* O* o0 6. 9 V.Ptabls 23 88* 39* 23.5 OA 393 373 36.3 *. 1" 6.7 1331 40.3 30.3 0D 0e - 54 569 10 sw 30* i8s' WD eD 91.A 270 36D* 3D e0 30*. OA 12J oA 5D 0e 3.3 0e 4*0 It Cd&a 0.9 23 2se 30s 0 1 39.7 31D 418. 5* 383 IO 37D 3eD DD o0 e0 0* a0 12 Cows 20 17.5 WD* 6.D0 * 3D.3 2en 337 J0s 22.8 O 23b.3 30* JO3 O0 2* 0* 6* 13 Tca O S.J 300 2.9 352 214 1534 m 3D0 2.2 0* 21.7 CA 44UJ O 0* 6D 4e 0 14 T dkaa 23 10.6 tSD 272 0*J 4*. XD* I0tD O OA 73 I3O DJ 45J 0* 0* 6* 6* I5 L.ab 0* 30.7 3.9 36.4 0 .LI 103 W2 3. 33.5 e0 27.9 te 4.4 0* 0.5 230 89.7 I6 G?asbm l30 30* 30*D IO 2* 8.9 2562 300 0* 5* 0.0 6.7 23* 4* 0.0 0A 33* 0o n7 Cqp 2* 2.0 5* I0* 0* * 10sn S3D O* 'A 0*D 23.7 0A 0* 0.0 3* OA 0* I3 hn mm,kima 2. 0o I0.0 I0* 0A S* 0 21.7 1. 2*0 O 6.7 0* O 0* 0* 35A es 19 S.,b- 2.0 0* 1 0A I5* 0* 7O.2 I0* 231 GO 5. 0O 6.7 l3* 23.3 e0 0* 473 O0 20 L t 2* 3te0 0* 23* 0* 0* t0* 25. 2* 3* 0* * 23* 23.3 0* 0o 35* 0* 21 Conamad 0* 30 1O* DA 0D 9.9 JOB 20* S* SO 0* 0* 0*D 40 o 0* 0 33* o0 22 Cutor ilaoad GO O 0* D3 0t2 9O tOD 25* t43 3J 0O I0* 0* 459 0* 0* 3* 30. 23 tbbw 4.9 273 O 2 e 2.4 Oa is 13.3 t92 113 2.94 4.2 27 2.4 323 0* 0o e 324. 49 24 PM%,ww 0*0 1220 ISO 2.I 0.3 2.4 6.3 t2.7 MA 9 O 0* 242 I0* 33* 0*J e0 teD 362 25 Wood 0*n 33J* 35* 30J O 10.5 2.4 19.1 37.8 23 0* 23A M0* 3.3 0* 0tt D DJ 26 Cak W 8.0 5* 0. Wm 3 3_* t 25 - 73 _ 23* o0 0o _ 320 - 27 Sl O0 332 0e 55.7 63 J 45.9 1D* 5D 30* 0*D 0 2eD 4* CA 0*D t0D OD* 23 Wad 1.7 23* 5D 42J 03 5.7 t0D 24.1 BA 26.? 0* 3I3 2D0 429 0o 6.8 319 029 29 cam 0* 38.2 23 ss.9 eo 72 5 30.6 I* 27.6 e0 7e IRA 409 0* eLI 11A *8n 30 ha. 2.0 0* 9.7 M?7 0.4 0.7 t0e 2en 6.9 2* O 0* 2?e 4* 0* 0* 44. S" 31 We 2* 63 SO 18* 02 5 M A0* 2eJ 2D 2. O 0* 0* 30D O 0* 30* 30* 32 PFbabuo 0*D 2.0 0A 29 O 3U4 4* 23D 2* 13 O 0* 30* 83 0D O* ISD V3* 33 Sem 2.0 12.9 5* 4* eD 33J 10D 25.7 3. 3U3 0.0 23.1 tDD 42.5 0D 0D ISO as 34 F*kw 0* 3.5 0* 5.9 SJ 32 5D 30* 2* 4 0*D 0* I0* 192 0* 0* IOD 30D 35 Ahbo. O 133t 0*D .3 0O 3JA I0 23* 0* 7.9 0* 5* 2 30r 0* n* 1323 464 36 I1 2D *0.2 0D 4.4 0 2. I8 19.9 2D 6A 0D 3.1 got AS- O* O I0*0 23.9 37. cq" 2D 20 SD 16 19 3.7 $A 210 2* 2.7 0* 2J 80* 15.7 0D OO 12.0 21A 38 Rawe 2D I8I 5* 34.9 O 5 8.0 23.5 0I 24A 0OD 14J 8O 233 0* o * 3O 0 29* 39 Lod 2.0 2.0 OA I0A OA 5J ID* 3* 2* 30 0D 123 O 2DD OA Oa O 19.1 4 21- 2.0 3.2 00 I0. 0o 4.6 I 23 2D 5* 0* 2J O 2DD 09 O* 0o 21.9 41 Tin 2D 2D OA 9sD o 33 In 23* 0* 50 0D 5tt 0* 2D0 0A 0O t0. 19.7 42 1uso 2_ 8.3 5D 32 O 2.4 I* 2A 2.0 6 0* 1.3 90 33.4 0* 0* tOD 0.0 43 T _aao 2.0 2* 5D SO 0e 3.4 3.1 2J 23 2.9 0* 0* IO 30* O Oa leD 23.5 44 Saow 2* I3J OD 25.0 OD 1A 13.7 JD* AGO 6O7 0D 0o JD sea 0D 0* 35* 35* 43 Ham a *Wb" 3* 8S.3 0lo 403 0* 13 3.6 2DJ 2.4 18.3 0D 25.3 36.7 37.9 0 Oo 33s 559 46 "air 1.6 a 9.2 g 06 - 366 W 3A m 0Di . 2D _ 0D _ 31. - 47 Feas8m 3. 8.1 92 37 OA 3.7 36.6 2DD 3IA 35s 03 9.7 25D 47n3 0D o 315 5DD 4 haak. 0D 0.3 oa S. 3S 63 5D 35.7 0* 4* OD 3.9 tOD 12. Oa [Oa 25* 292 P Fumap od *.h. v.1 ceodwm .__ ..... _ _ 71.7 4.3.9LS" 06.7 MA 76.3 78.3 *17 67.4 * A-up wk-y t_ ow tdW_m (5) 73 1G3 7* 024 tt 0) 9e 113 6I" 04 36.7 to Cmod km "ts Wat DB4-UNClAD SMARt dadb. Aw. ub. o yb. botb. g ad to4 5op a k d. 18 Malaysia - Korean import duties actually exceed 100 percent in several processing chains." The importance of these observations iF accentuated by the fact that average applied tariffs on processed comnmodities in the EC (our standardfor comparison) are about 6 percent (Laird and Yeats, 1987), w}ich is less than one third the average duty on Indonesian, Korean or Malaysian commodity imports. The second important point that emerges from Table 5 is that tariff escalation occurs in most commodity processing chains. In Japan, 89 percent of the processed commodities have higher average applied tariffs than do the primary stage components, and in several cases the spread in duties over a chain exceeds 30 percentage points (e.g., the cocoa and tobacco chains). Overall, the average difference in tariffs on primay and processed commodities is 16 percentage points or more for Thailand and Indonesia and about 12 percentage points for New Zealand. Japan's tariffs increase by 7 percentage points, which is approximately double that for the EC taiffs over the same processing chains.'2 The overall tariff difference for the Republic of Korea (about 9 percentage points) understates the true importance of escalation in most chains due to the perverse results in two chains -vegetables and tea, where relatively high tariffs are applied on the primary stage items. In both cases, nontariff protection may have been substituted for tariffs on the final stage good (see Table 6). While Table 5 indicates most countries' tariffs escalate sharply, other trade barriers may also contribute to Asia's import bias against processed commodities. Several Asian countries employ nontariff barriers, like quotas, licensing requirements, ot variable import levies to control imports. Thus, the presence (or absence) of NTBs over commodity processing chains should be examined to determine if these restrictions generally reinforce Asian tariffs. " The Japanese rates are 'applied' tariffs in that they are the average of the MFN or GSP duty actually paid on imports. The GSP rates may provide developing countries with important margins of preferences for many manufactured goods, but our analysis indicates they are not of major importance for primary or processed commodities. In some cases Australia, New Zealand and other countries that face MFN tarrifs may pay higher tariffs on a few products than indicated by the average rates shown in Table 5. 12 Yeats (198X p. 115) provides a cross-country comparison of the degree of escalation in 10 OECD countries' tariffs. Japan's tariffs escalate far more sharply than any of the other countries used as comparators. The increase in Japanese tariffs over commodity processing chains was about three times greater than that for the United States and at least six times greater than that for Norway, Sweden and Switzerland. ;' 19 Table 6 summarizes available information on nontariff .arriers applied by Asian countries on the 48 processing chains. The table indicates the specific types of nontariff measures that Indonesia, Japan, Korea, Malaysia and Thailand use to regulate primary and processed commodity imports. Since Australia, Hong Kong, New Zealand ard Singapore appear to make in"requent use of nontariff barriers on either primary or processed commodities these countries are not included in Table 6. Box 4 provides a somewhat different perspective by showing the share of processed commodity imports (in value terms) covered by nontariff measures. Table 6 shows that Asian NTBs are most highly concentrated in the foods and feeds and agricultural materials sectors. Japan uses eight different types of nontariff barriers to regulate food imports, including variable import levies on sugar, poultry znd pork and bovine meat and global quotas on fish and fruit. Related studies estimate that these Japanese nontariff barriers often have ad valorem equivalents greater than 100 percent (see Box 4). Other studies reach similar conclusions concerning the importance of Korea's NTBs on a number of food commodity imports (for example, see Anderson 1981). The key point that follows from these observations is that tariffs and nontariff barriers are of sufficient importance that both must be addressed in an Asian multilateral liberalization effori V. Suymma and Policy Recommendations While numerous studies have examined trade barrier escalaton and its implications in industral countries, very few related analyses have been undertaken for developing countries. This study undertook such an investigation in order to assess the priority to be given to the issue in intra-Asian trade. After compiling a comprehensive data base on Asian trade, tariff and NIB restrictions, this study examined the structure of these restrictions on primary and processed commodities. Most Asian countries' tariffs were found to be set at high levels and to incorporate grater escalation than industrial countries' import duties. Evidence was also presented that the escalation in the tariffs is often reinforced by nontariff restrictions on processed goods although it was not possible to draw firm conclusions on this point due to the lack of reliabltJ NTB ad valorem equivalents. 20 Tabl 6. Asira NondffMcea s Applied to Pimsfy and Fnal Stage Pocd Prod isn 48 Commodity Procesing Csin. I Meet of aow, .sbp. lamb, goat U. SP. GQ TQ VL IA 2 Swim U, SP. GQ VL IA CS 3 Pouay U U,SP VL IA IA 4 Fih fth, cbilod, U U I AQ IA. GQ IA IA S Shellah fresh, cild, fxen U, SP, GQ U [A. GQ IA, GQ IA IA 6 Whlat U GQ, Su CS U 7 Rice GQ. SM GQ, SM IA IA U Fruit U, Sp, GQ ST.GQ IQ, GQ IA IA CS CS U 9 Vegetabls U SP. SM GOQ,ORA IQ, im IA IA CS. IA CS- 10 Suar VL ET, IA U CS U U 11 Coffee SP, GQ GQ ET IA 12 Coeca 1, GQ ET Er 13 Tea U, SP, Q IA, GQ, SM IA ET, A CS 14 Tobao SEA CS CS IS Le-t%rr TQ CS 16 G _uesat U GQ IA CS CS 17 Capes Is , _hbnub.keh U IA 19 Sy abem Su IA U 20 L_med 21 Comm wod 22 C" oil seed 23 Rubber SP CS U 24 Pu ood IA CS 2S Wood IA 26 Codak 27 Silk SM IA IA U 2B Wool SP SP CS 29 Conoka SP CS 30 Jute U. SM OIM IA U 31 Flax IA 32 1hmpbat 33 SeI Er CS, U 34 Sulphr CS 35 Asbeda CS.,U U 36 Ion U. SP. Su CS. U 37 Copper 38 flausite CS 39 Lead IA 40 ZiDC 41 Tm 42 Manganese SP. SM CS. U 43 Tungam GQ 44 Siler IA 45 Hom & wlebeom U, GQ CS 46. Hair IA, GQ A 47 Fcasho [A. GQ GQ IA ET CS 48 Pletromasm TQ CS IA GQ - Gb6alQaud SM - Soat moolp TQ - Taniff Qua OIM =ehr Impost MO l ame [A l= mt Amrhemw SP = Selce lhdms VI - VaSPi - SIA Vt Solk Imo Atamy U Lic*c ST Semina Er - SETci- Tax CS - C Swcbmga Sec C_oputed fiee ekb Wodd Busk-UNCTAD SMART dat bae. No NITf an awarded bow rumay at prosd mpost proa a Autmrs. Now Zeskd at Sigapoe 21 Box 4 The Nature of Nontariff Protection for Processed Commodities Table 5 shows that some Asian e:ountries' tariffs are higher than those in OECD markets and frequently incorporate greater escalation. However, primary and processed commodities may also face nontariff barriers in Asian trade. GATT records show that while NTBs are applied infrequently to commodity imports by Australia, Hong Kong, New Zealand and Singapore (records for the Philippines are not available), they are widely used in countries like Japan, Korea and Malaysia. Share of prccessed commodity imports covered by nontariff barriers (9) Importing Country Foods & Ores & Agricultural Energy Feeds Metals Materials Products Indonesia 18.3 6.2 0.4 0.0 Japan '9.8 0.9 0.7 8.2 Rep. of Korea 18.4 0.0 22.2 85.4 Malaysia 23.2 22.3 36.5 52.5 Thailand 11.9 1.3 2.6 22.2 To illustrate this point, the above statistics show the percentage of tariff-line level processed commodities facing nontariff barriers in five Asian markets. In preparing these data, the 48 commodity processing chains wen aggregatd into four broad product groups. Processed foods are most often subject to NTBs - these restrictions cover approximately 50 percent of Japan's imports - but Korea, Malaysia and Thailand frequently apply nontariff measures to energy products (see Table 6 for information on the types of measures used in these markets). Although information on the trade effects of nontariff barriers is limited, several studies suggest Asian NTBs often convey very high levels of nominal protection. For example, Ss-on and Anderson (1982) estimate that Japan's NTBs on primary and processed foods often have nominal equivalents over 100 percent, while ECAFE (United Nations, 1982) reached similar conclusions concerning nontariff protection in other Asian countries. The message from these studies is that nontariff barriers, as well as tariffs, must be addressed in any multilateral liberalization effort. At a minimum, the NTBs should be bound" to ensure that they cannot be tightened to offset the effects of any tariff cuts that might be achieved in negotiations. 22 These findings strongly suggest that Asian trade bariers have an important restrictive effect on intra-Asian processed commodity trade. In short, this study demonstrates that trade barrier escalation is an important issue to be addressed in regional initiatives to liberalize trade barriers. The implications are that this problem should not be viewed as a pure South-North trade issue, which was the case in the Tokyo and Uruguay Round negotiations, but must be approached in a broader context. However, related lessons from previous MTNs have important itplications for future liberalizationefforts. One key point is to ensure that disproportionately high trade barrier cutFi are not made for unprocessed (as opposed to processed) commodities since this would work against expanded trade opportunities for processed products.13 A second message is that negotiations must deal jointly with tariffs and NTBs to ewsure that a reduction in one type of restriction is not offset by a further tightening in the other. As an illustraion, Laird and Yeats (1989) show that, over the 1966-86 period, industrial countries increasingly used nontariff barriers as a substitute for tariff protection. The results of this study call into question traditional explanations as to why trade bamier escalation exists. For example, Balassa (1968, p. 19S) indicates that trade barriers in the North increase with fabrication to 'discriminate against the processed export products of developing countries' and thereby protect the formers' producers from their more efficient counterparts in the South. However, the fact that Asian developing countries find it necessary to escalate their own trade barriers suggests that alternative or at least, less broad, explanations must be found. As a minimum, the South-North orientation suggested by Balassa must be dropped since tariff escalation to protect domestic industry against more efficient producers is not limited to developed countries. 'MTis conclusion follows from numerous studies that have analyzed the influence of trade barriers within a theoretical framework known as the "effective protection concept. Among other points, this concept holds that a proportionately greater reduction in trade barrers for production inputs as opposed to the final (manufactured) product may actually work against trade in the latter since these cuts would lower production costs and actually raise protection for the value added component of the manufacturing process. For a useful nontechnical discussion of the effective rate concept see Grubel (1971). For applications of the effective rate concept see Balassa, et. al. (1971) or Finger and Yeats (1976). 23 References Anderson, Kym (1981). Nolthi(ast Asian Agriculltural Protection in Historical and Comparative Perspective: The Case of South Korea, (anberra: Australia-Japan Research Center, May. Balassa, Bela (1965). 'Trade Liberalization and Revealed Comparative Advancage," The Manchester School of Economic and Social Studies, May. Balassa, Bela (1968). 'The Structure of Protection in Industrial Countries and its Effects on the Exports of Processed Goods from Developing Countries." in UNCTAD. The Ke_edy Round: Estimated Effects on Tariff Barriers. TD/6/Rev. 1, New York: United Nations. Balassa, Bela et. al. (1971). TheStructure of Protection in DeNelming Countries, Baltimore:Johns Hopkins Press. Basevi, G. (1966). "The United States Tariff Structure: Estimates of Effective Rates of Protection of United States Industries and Industrial Labor," Review of Ect tmics and Statistics, vol. 48. Commonwealth Secretariat (1982). Protectionism: Threat to International Ordej, London: Commnwealth Secretariat. Finger, J.M. and Alexander Yeats (1976). "Effective Protection by Transport Costs and Tariffs: A Comparison of Magnitudes," Ouarterlv Joumal of Economics, vol. 90, February. Grubel, Herbert (1971). "Effective Tariff Protection: A Non-Specialist Introduction to the Theory, Policy Implications, and Controversies," in Herbert G Grubel and Harry 0. Johnson (eds.), Effecdvo Tagff rotection, Geneva: GATT. Helleiner, G.K. and Douglas Welwood (1978). Raw Material Processing in Developing Countries and Reductions in the CanadiaTariff Ottawa: Economic Council of Canada. Johnson, Harry (1965). "The Theory of Tariff Structr with Special Reference to World Trade and Development, in H. Johnson and Peter Kenen, eds., Trade and Development, Geneva: Librairie Droz. Laird, Samuel and Alexarder Yeats (1987). "Empirical Evidence Concerning the Magnitude and Effects of Developing Country Tariff Escalation, The Developing Economira. vol. 25, June. Laird, Samuel and Alexander Yeats (1989). "Nontariff Barriers of Developed Countries, 1966-1986," Fijne and Development, vol. 26, March. LAird, Samuel and Alexander Yeats (1990). Ouantitative Methods for Trade Barrier Analysis, London: Macmillan Press. Roemer, M. (1979). Resource Based Industrialization in Develoning Countries: A Surve of the Literatur, Cambridge: Harvard Institute for International Development. Saxon, Eric and Kym Anderson (1982). Japanese Agricultural Protection in Historical Perspective, Canberra: Research School of Pacific Studies, Australian National University, July. Stern, Robert (1976). "Evaluating Alternative Formulae for Reducing Industrial Tariffs, Journal of World Trade L*w, January-February. 24 UNCTAD (1968). The Kennedy Round: Estimated Effects on Tariff Barriers, TD/6/Rev. 1, New York: United Nations. UNCTAD (1979). The Processing Before Export of Primara Commodities: Areas for Further Intemational Cooperation, TD/229/Supp. 2, Manila: UNCTAD. UNCTAD (1980). The Influence of Protectionism on Trade in Primara and Processed Commodities, TD/B/C.1/207/add. 1, Geneva: UNCTAD. UNCTAD (1991). Handbook of International Trade and Development Statistics, New York: United Nations. UNCTAD-World Bank (1989). A User's Manual for SMART, Washington: World Bank. United Nations Economic Commission for Asia and the Far East (1982). Intraregional Trade Proiections: Effective Protection and Income Distribution, vol. 2, Bangkok: United Nations. Walter, Ingo (1971). 'Nontariff Barriers and the Export Performance of Developing Countries,' American Economic Association Pagers and Proceedings, vol. 61, May. Walter, Ingo (1972). 'Nontariff Protection Among Industrial Countries: Some Preliminary Evidence," Economica Internazionale, vol. 55, May. World Bank (1981 and 1987). World Develooment Renort, New York: Oxford University Press for the World Bank. World Bank (1992). Global Economic Prospects 1992, Washington: World Bank. Yeats, Alexander (1977). 'Do International Transport Costs Increase with Fabrication? Some Empirical Evidence,' Oxford Economic Papers 29, November. Yeats, Alexander (1979). Trade Barriers Facing D2eveloDing Countries: Shigging and Commercial Policy Measure London: Macmillan Press. Yeats, Alexander (1984). 'On the Analysis of Tariff Escalation: Is there a Methodological Bias Against the Interests of Developing Countries?' Journal of Development Economics, vol. 15, Spring. Yeats, Alexander (1987). 'The Escalation of Trade Barriers,' in J. Michael Finger and Andnej Olechowski (eds.) The Urunuav Round: A Handbook for the Multilateral Trade Negotiations, Washington, World Bank. Y eats, Alexander (1991). 'Do Natural Resource-Based Industrialization Strategies Convey Important (Unrecognized) Price Benefits for Commodity-Exporting Developing Countries?' World Bank PRE Working Paper No. 580, Washington: World Bank, January. 25 APPENDIX 1 Elements of Primary Commodity Processing Chains Defined in Terns of Revision 1 of the Standard International Trade Classification System Table Al provides details on the commodity processing chains which form the basis for this study's empirical analysis. Each individual commodity is classified as falling in the primary, intermediate or final stage of a processing chain. To assist in clearly defining the nature of each item, its Standard International Trade Classification (SITC) Revision 1 number is also given.'4 A point to note is that chains defined in Table Al are based on the SfTC system and therefore may have certain limitations. Cne problem is that some of the SITC-based stages define products at too high a level of aggregation with the result that product composition may vary in ways that influence the empirical analysis. For example, the primary and processed stages of the fruit and vegetable chains may contain different proportions of (say) temperate and tropical products so they need not accurately represent a given (well-defined) commodity undergoing increased fabrication. A second problem concerns leakages from the chain. In these cases a given commodity experiences further processing, but is not used as a direct input into the next highest stage item. As a result, analysis of trade changes in a SITC-defined processing chain may understate the actual level of commodity processing and trade by developing countries. Finally, the SlTC system may contain some product groups that contain individual items which are at different levels of fabrication. As an example, the vegetable oil stages of the groundnut, linseed, soya bean, copra, and cotton seed chains do not distinguish between crude and refined oils although different levels of processing are involved. Several of the commodities listed in Table Al have end uses at the primary or intermediate etages of processing. For these items a processing chain analysis may show little progress in shifting exports to higher levels of fabrication. Soya beans are an example as the primary stage item is a feed product. Vegetables, fruits, fish, and shellfish are other processing chains where a strong consumer preference may exist for the fresh (unprocessed) stage of the product - a factor that would work against processing (preservation) in exporting countries. A final point to note is that there may be major differences in the number of stages that are identified for the processing stages listed in Table Al and this is often due to the nature of the SITC system. For example, several commodities like fruit, vegetables and fish have only a primary and final stage identified as SITC products. This contrasts with the wood (manufactures) chain where a primary stage, two intermediate, and two final stages can be identified. As a result of these differences in detail, it is very difficult to make cross-commodity comparisons of trade at similar levels of fabrication. "As an example, Table Al shows that the cocoa chain has three distinct stages with cocoa beans (SITC 072.1) representing the primary stage (unprocessed) product. Cocoa powder (SITC 072.2) and cocoa butter (072.3) are two items classified in the next highest stage of processing, while chocolate (SITC 073) represents an even higher level of processing activity. For products like wood manufactures it is possible to identify five different levels of commodity fabrication although some other chains, like petroleum, have only a primary and processed stage. 26 Table Al. Elements of the Wodd Rank's Commodity Processing Classification Schemc Processing Chain Prinary state product Sc intermwdiate oroduct(s) (SITC) Fina stage oducZ(s) (STC 1. FOODSTUFFS AND TOBACCO Pig meat Live wine (0013) Frcsh or fin pork (011.3) Presred pork (021.1, 013)a Poultry Live poultry (001.4) Fresh or fioe poujiy (OIIA, 011.81) Prepared or tinned mea (013)a Meat of Cows, sheep or goats Live cattle, sheep or goats Fresh or frzen beef or mutton (011.1, Met tinned or aoked (012 les 012.1,013)P (001.1,001.2) 00.2) Fish other than shellfish Fresh or frzen fisb (031.1) none identified Sated orpr-Aseved fis (031.2,032.01) Shellfish Fresh or fren sdelfish (031.3) none identified Pepard or presrved shelfish (031.02) Wheat Unmilkd wheat (041) Wheat mead or flour (046) Bread or biuits (04S.41) Rice Rice in husk or husked (042.1) none identified Rice gltad or pofished (042.2) Fmnit Fresh fruit (051) none identified Preeved fruit (053) Vegetables Fresh vegetables (054) none identified Prerved vegetables (055) Coffee Green or roasted beas (071.1) none identified Coffee exracs (0713) Cocoa Raw or roasted beam (072.1) cocoa powder and butter (072.2, 072.3) Cbocolate (073) Tea Tea (074.1) none identfified Tea extracts (099.02) Sugar Raw beet and cane sugar (061.1) Refined sua (061.2) Flavored -gar and ugar candy (062) Groundnuts Groundtat(221.1) none identified Groundwut oit (421.4) Copra Copra, excl. flour and meal (221.2) none identified Coconut oil (4223) Palm nuts and kernels Plm nuts and kemeb (221.3) none idaetified Palm kernel oil (422.4) Soya beans Soya bean excl. flour (221.4) none identified Soya bean oil (421.2) Unseed Linseed cxcl. flour (221.5) none identified Linmseed oil (421.2) Cotton seed Cotton seed excl. flour (221.6) none identified Cotton seed oil (4213) Castor seed Caster seed excl. flour (221.7) none identified Castor oil (422.5) Tobacco Unmanufactured tobwco (121) none identified Cigars, cigarettes, etc. (122) D. AGRICULTURAL MATERIALS Wood Rough logs (242.21,242.31) Log roughly equre (242.22,242.32)b Plywood and veneer (631.1,631.2)b Lumber sawn and planed Wood manufatures (632)c Cork Cork unworked (44.01) Cork simply worked (244.02) Cork manufacwu (633) Paper-pulpwood Pulpwood (242.1) Woodpulp (251.2,251.6,251.7,251.8) Paper (641.1 to 641.4, 641.7, 641.9)f, Paper atticles (641.1 to 6423. 642.9W Rubber Natural and synhetic bber (231.1, Unvuleanized tubber (621.01 to 621 .03)b ure, tubes and belts (629.1, 629A, 655.45)d 231.2) Vulcanized rubber (621.04 to 621.06p Rubber clothing (841.6)d Leather Hides and skins (211 less Bovineand sheep leather (611.3, 611.4, 611.91, 611.92)b Leather belting (612.1,612.2, 612.9)d 211.8, 211.9) Chamois tndpcahm kentlcthe(611.93, 611.94)c Leatherclothing (841.3,851.02) Feathers Feathers (291.96) none identified Feather goods (899.26, 899.92) Horn and whalebone Horn and whaebone (291.12) none identified Carved hon and whalebone (899.15) Hair Human hair (291.91) nmne identified Humian hair worked (899.94) Silk Raw silk (2613) Silk yam andt ded(6SI.1) SilkfMicka (653.1) lute Raw jute (264) Jute yar (6S1.9) lute fabrics (653.4) Wool Wool greasy (262.l)b Combed and rded wool (651.21,65122,651.25) Wool fabrics (653.21, 653.22)d ~2' Table Al. Elemens of the World Bank's Commodity Processing Classification Schema and Sources of Uni Vahle Information Processin, Chain Pimary sRe product (ST) Intermediate Wooducts) (S)lrM Findl staee producl(s) (STrC) We : degeaed (62.2, 262.6, Wool blankets (656.61)d 262.8)c Cotton Rawr cotton (263.1)b Conaonyarn(6513,651.4) Cotton librics (652.11 to 652.13, 652)d Coton combed ad carded (63.4)c Cotton blankets (656.62)d Flax, hemp and ramie Raw flax, hemp ad wmie (265.1 none identified Flax, hemrp ad raie yamn (651.5) to 2653) m. ORES, MINERALS AND METALS Phosphate Natural pbosphate (271.3) Pboswhic esktrs (512.63) Phosphoric acid (513.35)d Phosphate fenilizer (561.2)d Sulphur Sulphur (274.1) Ester and purified sulphur (512.61, 513.2) Sulphuric acid and conpounds (513.33, 513A2) Asbestos Crude Asbestors (276.4) none identified Asbet fiber and material (661.83, 663.8) Glass Sand excl. metal bearg (273.3) Glass in mss (664.11, 664.13, 664.3)b Safety or contruction gala (664.5 to 664.7)d Glass surface gromud (664)c Ghss ma ftcattres (664.91 to 664.93, 665.1, 665.2)d Iron ron ore (281.3) Pig iron (671.2)b Ion and steel plate (674.1 to 674.3)d Iron wire and tod (673)c io strip, rails, wires (67S, 676, 677)d Manganese Manganese ore (283.7) none idetified Fenno-manganese (674.1) Copper Copper oe (283.11,283.12) Copper unrfined (682.1 )b Copperbars, wire, plat (686) Copper refined (682.12, 682.13)c Nickel Nickel ore (283.21, 283.22) Unwrought nickd (683.1) Nickel bars and sheet (683) Bauxite Bauxite ore (283.3) Aluminum oxide (513.65)b Aluminum bars, plate, wire (684.2 less 684.24) Unwrought alumium (684.1) Lead Lead ores (283.4) Unwrought tead (685.1) Lead bars, plate, wire (685.2) Zinc Zinc ores (283.5) Unwrought zinc (686.1) Zinc bars, plate, wir (686.2) Tin Tin ores (283.6) Unwrought tin (687.1) Tin bars, plate, wire (687.2) Tungsten Tungsten ore (283.92) none identified Tungsten (689.41) Silver Silver ore (285.01) Unworked silver (681.1) Rolled silver (681.12) IV. PETROLEUM Petroeum Crude Petroleun (331.01) none identified Gasoline, kerosene, f6els (332.1t2. 332.3/4) a STC 013 (preserved mat) mnay contain other meats that do not strictly belong in the prcserving din. b Of the two products shown in this stage of the chain this item is less processed and may be a production c This item is considered to be the more highly processed of the twoin th mae stage of th. prcessing chi. d lbe items in tbis stage of the processing chain are classified as having sinilar levels of fabrication. 28 APPENDIX 2 Do Asian Countries Have a Comparative Advantage in Processing Natural Resource Products The previous analysis established three points relating to the priority that trade barrier escalation should receive in any Asian multilateral liberalization initiative. First, primary commodities constitute a relatively high share of many Asian countries' exports. Second, Asian imports are biased against processed products; in fact, processed Asian commodities fare considerably better in non-Asian markets than in intra-trade. Third, some important Asian countries, like Japan and Korea have relatively high tariffs that escalate sharply. A further important consideration that has not yet been addressed is whether Asian countries have a comparative advantage in processing primary commodities. Economists have employed the 'revealed* comparative advantage concept to ans-wer questions of this sort. Stated simply, the revealed comparative advantage (RCA) of country i in product j is measured by the item's share in the country's total exports relative to its share in world trade. That is, if x3 is the value of country i's exports of j, and Xo is the country's total exports, its revealed comparative advantage index is, RCAU = (qj1X)/(XX), where the w subscripts refer to world trade totals. The index RCA, has a simple interpretation. If it takes a value of less than unity (which indicates that the share of product j in i's exports is less than its share in world trade) this implies that the country has a revealed comparative disadvantage in the product. Similarly, if the index exceeds unity this implies that the country has a revealed comparative advantage in the item. (See Balassa (1965) for a discussion of the properties of this index)." '5One potential problem is that the measure can be influenced by other countries' trade barriers. If markets with which a given country has a strong tendency to trade have relatively high tariffs and NTBs, this could cause the RCA index to incorrectly indicate the lack of a revealed comparative advantage. Another potential problem is that the index can be biased by inappropriate economic policies. For example, a national strategy of import substitution could cause the RCA index to depart from its pattern under a "neutrald" trade regime. 29 Table A2. Asian Countries' Revealed Comparative Advantage in Processing Primary Commodities Procesed Commodity Group Exporter Foods and Feeds Agricultural Matenials Ores and Metals Energy Products Australia 0.75 0.55 1.91 0.42 China 1.00 2.16 0.42 0.46 India 0.56 3.39 0.43 0.99 Indonesia 1.21 5.51 0.44 2.23 Malaysia 3.07 1.24 0.37 0.52 Thailand 3.86 1.29 0.20 0.19 Taiwan, China 0.60 1.46 0.49 0.0S New Zealand 0.95 0.54 1.19 0.39 Philippines 4.24 1.44 0.88 0.38 Rep. of Korea 0.64 2.75 1.14 0.37 Brunei 0.00 0.01 0.01 1.62 Japan 0.12 0.65 1.19 0.12 Papua New Guinea 3.52 0.47 0.09 0.01 Singapore 0.71 0.38 0.20 6.40 Pakisan 0.28 6.17 0.00 0.11 Sri Lanka 0.50 1.21 0.06 0.42 Hong Kong 0.39 3.56 0.11 0.07 Bangladesh 0.25 2.65 0.12 0.62 Fji 5.96 2.46 0.01 0.25 Korea Dem. Rep. 0.89 0.29 4.96 0.54 Myarmar 0.59 0.19 0.16 0.21 Guam 0.55 0.17 0.S3 1.59 Maldives 15.04 0.01 0.10 0.01 Kampuchea, Dem. 0.30 0.10 0.03 0.00 Vanuatu 0.10 0.19 0.00 0.00 Solomon llands 11.19 0,01 0.04 0.00 Mongolia 0.04 3.34 0.46 0.00 Tonga 1.66 0.54 0.00 0.00 Nepal 0.82 0.32 0.01 0.00 Kiribati 0.28 0.00 0.00 0.00 Memo Item No. of countries With RCAs over unity 10 14 5 4 Source: Computations based on 1990 trade data drawn from United Nations Statistical Office COMTRADE records. Table Al identifies the specific processd commodities included in each of the above product groups. Table A2 shows 1990 revealed comparative advantage indices for 30 Asian countries' exports of four groups of processed commodities, namely, foods and feeds, agricultural materials, ores and metals, and energy products (see Table Al for a list of the processed commodities in each group). Overall, 25 of the 30 countries had a revealed comparative advantage in at least one of the four groups - a point that should indicate broad-based interest in escalation issues. Almost half of the countries (14 out of 30) have a revealed comparative advantage in processed agricultural materials, with 10 countries also having RCAs above unity for processed foods. Seven have a revealed comparative advantage in at least two of the four product groups, with Indonesia recording a comparative advantage in every group except processed ores and metals. Policy Research Working Paper Serles Contact Title Author Date for paper WPS1 151 Is Growth Bad forthe Environment? Charles van Marrewiik July 1993 J. Verbeek Pollution, Abatement, and Federick van der Ploeg 33935 Endogenous Growth Jos Verbeek WPS1152 Population, Health, and Nutrition: Denise Vaillancourt July 1993 0. Nadora Annuai Operational Review for Fiscal Stacye Brown 31091 1992 and Others WPS1 153 North American Frea Trade Aiberto Musalem July 1993 P. Infanto Agreement: Issues on Trade In Dimitri Vittas 37664 Financial Services for Mexico Ashl DemirgOg-Kunt WPS1 154 Options for Pension Reform In Tunisia Dimitri Vitas July 1993 P. Infante 37664 WPS1 155 The Regulation and Structure of Martin F. Grace July 1993 P. Infant. Nonlife Insurance In the United Michael M. Barth 37664 States WPS1 156 Tropical Timber Trade Policies: What Panayotis N. Varangis July 1993 D. Gustafson Impact Will Eco-Labeling Have? Carios A. Primo Braga 33714 Kenji Takeuchi WPS1157 Intertemporal and Interspatial Sultan Ahmad July 1993 E. O-Reilly- Comparisons of Income: The Meaning Campbed of Relative Prices 33707 WPSI 158 Population Growth, Externalities, Nancy Birdsall July 1993 E. Homsby and Poverty Charles Griffin 35742 WPS1 159 Stock Market Development and Ashl DemirgOg-Kunt July 1993 P. Sintim- Financial Intermediary Growth: Ross Levine Aboagye A Research Agenda 38526 WPS1160 Equity and Bond Flows to Asia and Punam Chuhan July 1993 R. Vo and Latin America: The Role of Global Stjn Claessens 31047 and Country Factors Nlandu Mamingi WPS1 161 Increasing Women's Participation In Molly Maguire Teas July 1993 L Maningas the Primary School Teaching Force and 80380 Teacher Training in Nepal WPS1162 The Slovenian Labor Market In Milan Vodopivec July 1993 S. Moussa Transition: Issues and Lessons Samo Hribar-Milic 39019 Leamed WPS1 163 Domestic Distortions and James E. Anderson July 1993 D. Gustafson intemational Trade J. Peter Neary 33714 Policy Research Working Paper Series Contact Title Author Date for paper WPS1 164 Power, Distortions, Revolt, and Hans P. Binswanger July 1993 H. Binswanger Reform in Agricuitural Land Relations Klaus Deininger 31871 Gershon Feder WPS1 165 Social Costs of the Transition to Branko Milanovic August 1993 R. Martin Capitalism: Poland, 1990-91 39026 WPS1 166 The Behavhiorpf Russian Firms in Simon Commander August 1993 0. del Cid 1992: Evidence from a Survey Leonid Liberman 35195 Cecilia Ugaz Ruslan Yemtsov WPS1 167 Unemployment and Labor Market Simon Commander August 1993 0. del Cid Dynamics in Russia Leonid Liberman 35195 Ruslan Yemtsov WPS1 168 How Macroeconomic Projections Rashid Faruqee August 1993 N. Tannan in Policy Framework Papers for the 34581 Africa Region Compare with Outcomes WPS1 169 Costs and Benefits of Debt and Eduardo Fernandez-Arias August 1993 R. Vo Debt Service Reduction 33722 WPS1 170 Job Search by Employed Workers: Avner Bar-Ilan August 1993 D. Ballantyne The Effects of Restrictions Anat Levy 37947 WPS1 171 Finance and Its Reform: Beyond Gerard Caprio, Jr. August 1993 P. Sintim- Laissez-Faire Lawrence H. Summers Aboagye 38526 WPS1172 Liberalizing Indian Agriculture: Garry Pursell September 1993 D. Ballantyne An Agenda for Reform Ashok Gulati 37947 WPS1173 Morocco's Free Trade Agreement with Thomas F. Rutherford September 1993 N. Artis the European Community: E. E. Rutstrom 38010 A Quantitative Assessment David Tarr WPSI 174 Asian Trade Barriers Against Primary Raed Safadi September 1993 J. Jacobson and Processed Commodities Alexander Yeats 33710